p.2 #1 · F-Stops vs T-Stops for modern photo lenses?
'While f-stops are standard, T-stops (transmission stops) are used to measure the actual light passing through, which can vary by 10-40% due to lens glass.'
Is 10-40% close enough for you? OK, but we are forced to accept it, not by choice. A better statistic is available but it's not being divulged. Would you think lens makers don't know their lenses' T-stops?
F-stops are formulaic entities. T-stops measure one very important aspect of lens performance. Don't you find it interesting that - for stills lenses, but not cine lenses - the industry could choose to not report such a valuable performance measure?
These are very interesting curves - thanks for the link. They confirm a long-held suspicion of mine: the transmission curves differ at most by a few percent, and they are "smooth" by which I mean that a simple (and small) color temperature adjustment will bring these into close enough alignment that I don't think you could tell the difference in an image. Someday when I have time, I'll play more in photoshop, but I have a feeling that bumping the B channel by 1 and the G channel by 0.5 in the mid tones (which is just over a 1% change for a mid-range value of 128) will be barely perceivable in color. Maybe I'm wrong! But I'm kind of skeptical that most people can see 1% changes (there's a reason that 256 was chosen for the max count in digital imaging: we just can't reliably perceive a single count difference).
Modern optical coatings are so good that I think the main origin in color differences between lenses has little to do with coatings and much more to do with the angle of incidence of rays onto the sensor from the exit pupil of the lens, which does vary quite a bit from lens to lens (together with the angular dependence of the color filters on the sensor)_.
p.2 #3 · F-Stops vs T-Stops for modern photo lenses?
gdanmitchell wrote:
IT is almost like we're looking for things to get upset about.
We're all photographers here, right? How many of us have ever encountered a real world problem because f-stops and t-stops tell us different things about our lenses?
IN what are apparently the worst, outlier situations... when you shoot wide open in the most marginal light you might be 1/3 stop underexposed. That's basically background noise...
I have no objection to manufacturers adding t-stop info to their spec sheets, but...
p.2 #4 · F-Stops vs T-Stops for modern photo lenses?
philip_pj wrote:
F-stops are formulaic entities. T-stops measure one very important aspect of lens performance. Don't you find it interesting that - for stills lenses, but not cine lenses - the industry could choose to not report such a valuable performance measure?
To be fair, transmission is probably more of an issue in the cine world. Videographers want to be able to change focal lengths without exposure or color changing, which would make editing much harder to create a coherent look. Differences in photos are usually much more forgiving, even if the photos are displayed side by side. Also, widely available raw video is fairly new compared to raw photos, so videographers had to get it right in camera more than photographers.
It would definitely be nice to have, though. Sigma and Zeiss put T-Stops on their cinema lenses, but not their photo lenses.
p.2 #5 · F-Stops vs T-Stops for modern photo lenses?
I suspect that many still lens manufacturers would come out looking bad to inept if they published T stops, especially companies we tend to hold in high regard like Voigtländer.
p.2 #6 · F-Stops vs T-Stops for modern photo lenses?
First, a question: What source are people using to find t-stop values for various photographic lenses? (I found mention of a DXO lens database that might include the information, but the link was dead.)
- - -
gdanmitchell wrote:
IT is almost like we're looking for things to get upset about….
chez replied:
… it turns out that the t-stop is not much different than the much cheaper and smaller 1.8 lens.
You are presuming that the issue with t-stops affects only the f/1.4 lens and not the f/1.8 lens. It affects all lenses to some extent.
- - -
DWOfPaul wrote:
Just to be clear, I am not upset about anything, just curious. To me, it starts to get interesting when you think about cases such as f1.4 vs f1.2 primes. Would it really be worth the increased size and cost of f1.2 over f1.4 if you're not getting a .5 stop more light gathering from f1.2?
An analogy if you will…
This is a different example than I posted originally, but it might be a better one.
“EVs never actually get the EPA listed range.The range for that EV is listed as 300 miles, but in real driving it is always less. I might as well get an EV that is listed as having 270 miles.”
But if EVs never actually get the listed EPA range figure… the 270 miles range EV is going to get less than that, too!
Something similar is the case with lenses. ALL lenses “lose” some light as it passes through lens elements — some more and some less. (More elements generally will lower the t-stop, as will certain other factors.)
If you are concerned that a f/1.4 lens loses some light and it might act like a f/1.5 lens… if you decide you might as well get a f/1.8 lens it is also subject to some light transmission loss and maybe it acts like a f/1.9 or f/2 lens.
See the problem?
BTW, there are no “perfect lenses.” Every lens “absorbs” light to some degree. :-)
- - -
RustyRus wrote:
If you don't care- Bow out of the conversation-
p.2 #7 · F-Stops vs T-Stops for modern photo lenses?
gdanmitchell wrote:
If you are concerned that a f/1.4 lens loses some light and it might act like a f/1.5 lens… if you decide you might as well get a f/1.8 lens it is also subject to some light transmission loss and maybe it acts like a f/1.9 or f/2 lens.
See the problem?
BTW, there are no “perfect lenses.” Every lens “absorbs” light to some degree. :-)
Before diving into this topic, that was my thinking too. Each lens is going to lose a similar amount of light since no lens has zero loss. But take a look at the comparison I posted earlier in this thread between the Sony 50mm f1.4 ZA and the Sony 50mm f1.8 ZA: https://www.fredmiranda.com/forum/topic/1939263/0#17017995
The total T-Stop difference between them is not 2/3 of a stop, but 1/3 of a stop, and if you take into account vignetting, the 55mm pulls ahead by about 2/3 of a stop in the corners. So, for low light photography, there is an argument that the smaller, lighter, and cheaper 55mm lens is actually better.
p.2 #8 · F-Stops vs T-Stops for modern photo lenses?
RoamingScott wrote:
I suspect that many still lens manufacturers would come out looking bad to inept if they published T stops, especially companies we tend to hold in high regard like Voigtländer.
Humm, you're making me want to test my test my CV lenses, especially since I often put the 35mm f2 APO in my bag for low light use. Based on your experience, do you think it would be fair to test it at f4 and smaller against the Nikon 24-120mm f4, or would I be better off testing it against the Zeiss 35mm f2? (Ideally, it should probably be tested against the Nikon 35mm f1.8, but I don't have that lens).
p.2 #9 · F-Stops vs T-Stops for modern photo lenses?
Well the thesis is that heavy vignetting would reduce the overall transmission, and the APO clears up a lot by F4. It would be interesting but likely not actionable. I’m more interested in how lenses act close to wide open in relation to each other.
DWOfPaul wrote:
Humm, you're making me want to test my test my CV lenses, especially since I often put the 35mm f2 APO in my bag for low light use. Based on your experience, do you think it would be fair to test it at f4 and smaller against the Nikon 24-120mm f4, or would I be better off testing it against the Zeiss 35mm f2? (Ideally, it should probably be tested against the Nikon 35mm f1.8, but I don't have that lens).
p.2 #10 · F-Stops vs T-Stops for modern photo lenses?
DWOfPaul wrote:
Before diving into this topic, that was my thinking too. Each lens is going to lose a similar amount of light since no lens has zero loss. But take a look at the comparison I posted earlier in this thread between the Sony 50mm f1.4 ZA and the Sony 50mm f1.8 ZA: https://www.fredmiranda.com/forum/topic/1939263/0#17017995
The total T-Stop difference between them is not 2/3 of a stop, but 1/3 of a stop, and if you take into account vignetting, the 55mm pulls ahead by about 2/3 of a stop in the corners. So, for low light photography, there is an argument that the smaller, lighter, and cheaper 55mm lens is actually better. ...Show more →
A few things:
1. I’m a fan of small, light lenses with smaller maximum apertures for many things, but mostly because they perform well, are less obtrusive, along with being lighter and smaller.
2. The fact that the f-stop doesn’t equal the t-stop on lenses is not a new thing. It has essentially always been that way. (It always will be, too. No camera lens has perfect 100% transmission.)
3. More striking differences between specific pairs of lenses (such as in your example) may be of interest to people considering those two lenses, of course, but they represent extreme cases, not the norm. As we read at your link, “ While for the most part the differences are not notable, there are a few that are.”
p.2 #13 · F-Stops vs T-Stops for modern photo lenses?
It's a bit of a mess in the stills world, because the big industry players want it that way. The numerical relationships are not arithmetic, hence the best you find in the literature is modified by words like 'roughly' and 'approximately'.
I encountered this when looking into f1.2 lenses. You would think f1.2 was more or less exactly a half stop between f1.0 and f1.4, but many sources put out gibberish this way or similar:
'F1.2 vs F1.0: f/1.0 is a full stop faster than f/1.4, meaning f/1.2 sits between them, making f/1.2 about 1/2 to 2/3 of a stop slower than f/1.0.'
'f/1.2 is roughly a half-stop (technically about 0.47 to 0.5 stops) brighter than an f/1.4 lens, allowing in roughly 36–50% more light. However, it is not a half-stop slower than f/1.0; it is actually about one full stop slower than an f/1.0 lens.'
We should want demystification, not more obfuscation. Hence the advocacy for the adoption of T-stops for all lenses. Besides giving you an accurate figure of lens transmission performance, it actually means something. It's categorically different.
If you want to take lenses seriously as optical instruments, you don't want to see companies throwing their dirty laundry in the waste paper basket, like they do with focus breathing and much more besides, like entrance pupil data etc.
'Authentic cine lens makers generally do not use f-stop markings on their lenses, opting instead for T-stops (Transmission stops).' It's not hard to do..but it won't happen until video drags them kicking and screaming into the modern world, and even then it's just a maybe.
It has to be someone big. If Thypoch (which makes the same lenses in cine and stills) put T1.5 on its stills lenses, people would simply see it as yet another feature they can moan about, along with the unusually spaced aperture markings etc. People are addicted to what they think is 'normal' and they (often vociferously) resist change, at every step forward.
p.2 #14 · F-Stops vs T-Stops for modern photo lenses?
I thought most cine is done with the PL mount, which is quite different from the mounts used for still/hybrid consumer cameras.
I doubt that some cheap Chinese lenses will easily convince Sony/Canon/Nikon to use T-stops on lenses for still use. Just mapping the electronic aperture controls (1/3 stop increments) and EXIF data would be a nightmare since the f-stop is what matters for DOF, not the T-stop. When I set a lens at f/8 for stills I do it for the DOF and optical properties of the lens, not the amount of light to worry about effective T9 exposure. A slight change of shutter speed or ISO can compensate.
p.2 #15 · F-Stops vs T-Stops for modern photo lenses?
philip_pj wrote:
It's a bit of a mess in the stills world, because the big industry players want it that way. The numerical relationships are not arithmetic, hence the best you find in the literature is modified by words like 'roughly' and 'approximately'.
I encountered this when looking into f1.2 lenses. You would think f1.2 was more or less exactly a half stop between f1.0 and f1.4, but many sources put out gibberish this way or similar:
'F1.2 vs F1.0: f/1.0 is a full stop faster than f/1.4, meaning f/1.2 sits between them, making f/1.2 about 1/2 to 2/3 of a stop slower than f/1.0.'
'f/1.2 is roughly a half-stop (technically about 0.47 to 0.5 stops) brighter than an f/1.4 lens, allowing in roughly 36–50% more light. However, it is not a half-stop slower than f/1.0; it is actually about one full stop slower than an f/1.0 lens.'
We should want demystification, not more obfuscation. Hence the advocacy for the adoption of T-stops for all lenses. Besides giving you an accurate figure of lens transmission performance, it actually means something. It's categorically different.
If you want to take lenses seriously as optical instruments, you don't want to see companies throwing their dirty laundry in the waste paper basket, like they do with focus breathing and much more besides, like entrance pupil data etc.
'Authentic cine lens makers generally do not use f-stop markings on their lenses, opting instead for T-stops (Transmission stops).' It's not hard to do..but it won't happen until video drags them kicking and screaming into the modern world, and even then it's just a maybe.
It has to be someone big. If Thypoch (which makes the same lenses in cine and stills) put T1.5 on its stills lenses, people would simply see it as yet another feature they can moan about, along with the unusually spaced aperture markings etc. People are addicted to what they think is 'normal' and they (often vociferously) resist change, at every step forward. ...Show more →
I don't think this is an "industry" issue. That's just how f-stops work.
First off, an f-stop doesn'claim to be a measurement of light transmission. It is a way of describing the relationship between the nominal focal length and the size of the aperture.
Because of the non-linear nature of the f-stop scale, a value (like f/1.2) that appears to be midway between two whole stops (like f/1.0 and f/1.4) is actually too small to be halfway in between. The aperture would have to actually be a bit larger than f/1.2. (Just look at the whole stops to see how this would work. The aperture midway between f/2 and f/4 is not f/3, it is f/2.8.)
The need for t-stops in the movie industry is not about not "hiding dirty laundry" or "tak(ing) lenses seriously as optical instruments." It is about a different set of needs that historically exist in film that don't exist in the same way in still photography.
A thought experiment: Those of you who own more than one lens, particularly if you own a combination of zooms and primes and/or OEM and third party lenses, have always been using lenses with varying transmission levels. Has this caused a real-world problem for you in the past?
Again, I have no objection to lens manufacturers or reviewers sharing transmission data about lenses, but for typical still photographers it really isn't a critical thing.
p.2 #16 · F-Stops vs T-Stops for modern photo lenses?
gdanmitchell wrote:
A thought experiment: Those of you who own more than one lens, particularly if you own a combination of zooms and primes and/or OEM and third party lenses, have always been using lenses with varying transmission levels. Has this caused a real-world problem for you in the past?
It did in the film era. Exposures were inconsistent between apertures and typically darker wide open. We learned to mentally compensate that first f-stop or so.
p.2 #17 · F-Stops vs T-Stops for modern photo lenses?
Best practice should always be implemented regardless of one's need for it, because it is the right thing to do. We are philosophically at the extreme opposite ends of the issue, obviously.
I don't like corporate secrecy, dishonesty or under-reporting of (in this case) lens performance, but I understand others have a (much) greater tolerance of these things. It's one reason to highlight them here, in fact, so readers can see both sides of the debates. cheers.
p.2 #18 · F-Stops vs T-Stops for modern photo lenses?
gdanmitchell wrote:
I don't think this is an "industry" issue. That's just how f-stops work.
First off, an f-stop doesn'claim to be a measurement of light transmission. It is a way of describing the relationship between the nominal focal length and the size of the aperture.
Because of the non-linear nature of the f-stop scale, a value (like f/1.2) that appears to be midway between two whole stops (like f/1.0 and f/1.4) is actually too small to be halfway in between. The aperture would have to actually be a bit larger than f/1.2. (Just look at the whole stops to see how this would work. The aperture midway between f/2 and f/4 is not f/3, it is f/2.8.)
The need for t-stops in the movie industry is not about not "hiding dirty laundry" or "tak(ing) lenses seriously as optical instruments." It is about a different set of needs that historically exist in film that don't exist in the same way in still photography.
A thought experiment: Those of you who own more than one lens, particularly if you own a combination of zooms and primes and/or OEM and third party lenses, have always been using lenses with varying transmission levels. Has this caused a real-world problem for you in the past?
Again, I have no objection to lens manufacturers or reviewers sharing transmission data about lenses, but for typical still photographers it really isn't a critical thing. ...Show more →
Ignorance is bliss. Many people buy a 1.4 lens not for its bokeh, but for its extra speed for dim situations. How would they feel if they discovered that 1.4 lens they spent big bucks on really is not that much better in low light than a much cheaper and smaller 1.8 lens? Not knowing this fact…everyone is happy with the 1.4 lens.
True world example. Everyone was happy buying the latest bag of potato chips until it was made public that there is 15% less chips in the bag due to shrinkflation…now they are upset.
p.2 #19 · F-Stops vs T-Stops for modern photo lenses?
EB-1 wrote:
It did in the film era. Exposures were inconsistent between apertures and typically darker wide open. We learned to mentally compensate that first f-stop or so.
EBH
Yep, shooting film with an external meter you quickly got to know your lenses and compensated the exposure for their differences.
p.2 #20 · F-Stops vs T-Stops for modern photo lenses?
philip_pj wrote:
Best practice should always be implemented regardless of one's need for it, because it is the right thing to do. We are philosophically at the extreme opposite ends of the issue, obviously.
I don't like corporate secrecy, dishonesty or under-reporting of (in this case) lens performance, but I understand others have a (much) greater tolerance of these things. It's one reason to highlight them here, in fact, so readers can see both sides of the debates. cheers.
I don’t like conspiracy theories, and this thread is leaning hard into one.
It's one reason to highlight them here, in fact, so readers can see both sides of the debates. cheers.